Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[GavJ]

Oh wow I didn't know that it didn't already do the reaction chains correctly. That is a good one.

I guess also along that same vein you could do the correct AMOUNTS as well as an intermediate step.

1) How much iron and whatever does this civ have across its sites, or broken down by site, compared to other civs or sites or compared to average map area?
2) (somehow) How much iron is use in steel production? How much is used in anvil production, etc.? Raws would be nice, could be very simple, similar to the [FREQUENCY:100] system in place for minerals. As in,  the game adds up everything's frequencies and determines actual amounts proportionally. Anvils at FREQUENCY:30 and steel at FREQUENCY:50, and iron itself at FREQUENCY:20 perhaps -- for every material look up all the things it can be plus its own number, and divide up how much of it is used or left native accordingly. In this case 200 iron would be split up 100 for steel making, 60 for anvil making, 40 left as iron.
3) And so on down the whole reaction trees automatically.
4) The intermediate value of this is that the caravans could for now use those numbers to decide how much of each thing to bring proportionally, which would be cool.

Yes, this would indeed require stone anvils, etc.

To make those proportions entity-specific, you could do [FREQUENCY:30] = all entities if you want, but optionally also [FREQUENCY:MOUNTAIN:20] [FREQUENCY:PLAINS:40] to split up if you want (those are the raw codes for dwarves and humans, anybody not familiar with modding)



These go in the existing raws for any finished goods or materials that aren't naturally occurring. For instance, hematite is naturally occurring, so FREQUENCY in that refers to how often it shows up in the rocks. Whereas iron is not naturally occurring, so FREQUENCY is interpreted as above for trade purposes.

[GoblinCookie]

My major concern with this is actually not the "controversial part" (I still think that's silly, but since you present it as an extra separate caravan, it implies you could turn it off if you want). My problem is the complete un-customizability of the rest of it and how much it would ruin modders.

It would only be uncustomisable and ruining to modders if the basic variables it functions on (rather than the system itself) are hardcoded.  That would be a possible bad design design decision on the part of the developers, but not anything that follows from what I wrote as far as I am aware.

You didn't mention how the settlements get all the other resources besides minerals, and the fact that they are all different and unique is a trap that's drawing you toward a bad place. An abstract cost to mine system in the raws accounts for anything and everything in one way, even things you haven't predicted or in heavily modified games, no problem. But your system where you actually start to simulate gameplay with hard-coded algorithms industry-by-industry does not.

I did not mention other  because they all could operate according to the standard model of production as far as I am aware, the main difference with miners is they can gradually increase the number of z-levels of resources the fortress can get.  The dwarves are given jobs producing whatever it is, that creates a demand for the necessary materials and the tools which are treated as reserved stockpile.

The game simulates production based upon the raw numbers of dwarves working in different occupations and said dwarves produce a certain amount each month.

Resources other than metal are processed as a limitation on the amount of dwarves that can be employed in a given profession.  The number of trees limits the number of woodcutters, the amount of soil limits the number of farmers.  The AI for simplicity's sake does not engage in environmental destruction, it only harvest resources sustainability.

Making woodcutters demands axes, but the total number of woodcutters is limited by the total number of trees.

As I hinted at in my previous post, humans have all kinds of crazy things that motivate them and cause us all to play the game wildly differently. The more you simulate NPCs explicitly, and uncustomizably, the more you just alienate me if I'm not playing YOUR specific type of fortress, because nobody else in my civilization will be abiding by my special needs or desires or mods. If you can mod the abstracted system, then you can just mod these things in and everybody is happy. If it's carefully simulated in core code, you can't, and only people who play exactly like the code suggests they should get the correct experience.

That simply is not true.  The system of supply and demand keep track of what the human player's settlement produces, imports and exports and it is automatically fitted into the system.  It is a fluid system that will take into account player differences. 

Wool, logs, meat, fish, silk webs, cheese, plants and seeds, leather, tree fruits, water, sand, clay, etc. etc.  None of which you've accounted for yet, each needs to have its own set of (FPS-lowering) detailed fort-by-fort simulations of how much they've dug or cut or sheared and who's doing it and when and how often, and you'd have to keep numbers on all the trees and plants and animals by species in every settlement and how they relate, and blah blah.

Not so because nobody is actually doing to be doing those things.  All that is really going on is that a certain number of persons of profession X exist and they are producing a certain amount of goods.  No actual trees get cut down, no actual animals get sheared.  Those are simply limitations on how many persons can be assigned to those professions in the settlement. 

The only actual numbers that are changed are the z-levels and the amount of actual finite resources (stone, metals, gold, gems etc), these are subtracted accordingly as those resources are 'produced' by miners.  This is just numbers, any decent computer adjusts thousands of numbers constantly without a blink. 

And what this does is, well A) slow down your game by quite a lot, but more importantly B) essentially lock in dwarves as the only reasonable species to play, by assuming their whole culture and industrial model and hard-coding it where we can't change it anymore. So then what do you do if modders add their own new materials that you haven't accounted for yet, or remove some, or completely change how stuff is mined (for instance, by requiring that wood by sawn and dried before use, using custom workshops)? None of this is possible anymore when the economy is specially simulated by NPCs versus abstract goods. At best it makes NPCS act completely nonsensically, at worst it crashes the game. You'd be undoing large amounts of flexibility that Toady has built into the game in the past.

The materials demanded by different professions can be written in a text file.  This means that modders can quite easily alter the materials demanded by the NPC simulated production so it is in line with their modified production.

Exactly the same thing is true with the demands of goods of different civilians of different species.

Even more problematic, what do you do if I'm not playing as a dwarf, but as a custom species that screws up your hard-coded demands? How does your system would work if I'm using a custom species which does not eat (no farming, no demand or trade in food) or drink (no booze) or have emotions (need to be able to mod out again irrational nationalistic emotions even if you put them in as an option), and is amphibious, and lays eggs, is naked all the time, has 8 legs and all custom clothing, etc.?

Mine would work just fine against such a challenge. Notice that nowhere in my entire suggestion is there a single specific resource or assumption about what species is playing or what their society is like, because you can completely modify all demands and all production uniquely for each entity. You can easily make spiders that demand bones and produce silk for trade and use custom-modded nectar and clay to build their homes, and that make armor out of shells, and that prize kaolinite as their version of gold/jewels but find actual gold/jewels/coins useless biped toys. No problem-o, just whip up some custom raws. The complexity of my plan is almost exclusively in the complexity of the raw tags and other customizability I suggested, not the actual economics, which are like maybe 25% of the effort. The reason is because of this stuff I'm describing now.

I used dwarves as an example but did not have time to get into the fact that different races/species have both separate demands and separate professions activated or deactivated.  Elf woodcutters I envision would be customized to produce special elf woods and elf wood using professions only demand special elf woods, while producing special elf wood items.  Elf consumers have no demand for normal wood items at any level of affluence.

If you want to make your first creature work in my system, you would go into the files and delete all reference to ANY_FOOD, ANY_DRINK or specific food/drinks from all levels of consumer affluence.  You would then change the available z-levels to eliminate the surface level and above, while also removing the bit that prevents counting of underwater resources.

The system works according to it's hard-coded economic logic, but the system is built on soft-coded data, such that it will behave flexibly according to what makes sense given that data as it behaves flexibly according to the game world itself.

Raw greed for wealth (gems, precious metal, coins, and high priced commodities in general), does actually sound like an interesting mechanic to me.  It would drive conflict/motivations once the basic need demand are met.  Plus it fits with the whole greedy dwarf mythology.  Rating the demands by level of importance versus affluence seems like it might work.  Like food is always high importance, but caviar and brie are luxury items that forts only prioritize as they reach a high level of wealth.

That was just my original money idea that kind of got swept under the carpet in the political argument about subsidizing caravans. 

The beauty of this system is that it simulates currency without hard-coding it into the game.  You want a society that uses stone as money, then all you do is replace coins with ANY_STONE or a particular stone if you prefer.  When it has met all it's demands, it will sell it's surplus for stone and this in turn will completely alter the economy, so that stone which at present is superabundant for everyone meaning that it is essentially untradable suddenly becomes something you can sell to caravans in vast quantities.   

If you want to get rid of money altogether, you simply set the highest affluence goods for a race/species to be finite.

[GavJ]

In that case, GoblinCookie, if you aren't actually simulating industrial production itself for ANYTHING other than mining, then I'm afraid I don't really see much or any difference between our proposals?

The only things that seemed particularly meaningfully different were the lack of mention of any RAWs (which you say you do intend, though), and what appeared to be a different more detailed simulation of the production system (but apparently not). So they're almost identical... which is good I suppose.

The beauty of this system is that it simulates currency without hard-coding it into the game.  You want a society that uses stone as money, then all you do is replace coins with ANY_STONE or a particular stone if you prefer.

Although this will work as you say, I'm not really sure why you're describing it as a special subsection of your idea or whatever. In a proper economic model, ALL goods (or nearly/most) should have infinite demand to them. Although the demand tapers down the more of an overabundance of something you have. I.e. your 2,000th pound of rice you're willing to pay less for than your 50th pound of rice.

But as long as a good has 1) physical space for you to store it (which is easy in DF) and 2) any conceivable possible future use whatsoever,
then it will still have continued demand. In the example of rice, dwarves would definitely have a potential use for much more rice than they eat a year: sieges.

Therefore you should already have dozens of goods with "infinite" (asymptotically tapering) demand, and thus these would all already do what you want your money to do on their own, without having to write any special lines of code.

By the way money also has that same feature in real life -- additional amounts of it are progressively worth less to you the more you have of it, just like every other commodity (this is a psychological fact). Money is just a regular commodity (outside of certain bank shenanigans, etc. which don't bother us because we don't have banks). So yeah sure I'd put an infinite (tapering) demand on ti and it would work this way, but so should a whole bunch of other things. And without any need for special code.




(I'm a little bit concerned about what you MEAN by demand if not this -- infinite but tapering supply for everything except hugely bulky goods)

[GavJ]

Okay sorry, I should be more explicit and clear.  This is relevant to both/all our plans, and I've been dancing around it too long and procrastinating, so I'll just dive in.

Supply and Demand Mechanics
Okay so first, terms:
"Demand" is a willingness to pay for something. However, this is not a constant amount for a good, but changes along a curve, seen below. Supply is the willingness to produce something at a given price point, and also changes along a curve. For our purposes, think of the demand curve as "price as a function of quantity" and think of supply as "quantity as a function of price."



So this is the basic vanilla situation. Demand is a curve, as shown, that depicts how much you're willing to pay for something at any/all possible quantities.  Usually in real life, it's not a straight line like that, but would asymptote as it approaches the X axis (though it certainly CAN intersect), but don't worry about that too much just yet. Price is effectively how much producers will produce when tempted with different price points. More $$ encourages them to produce more to take advantage of the greater profits.

The price something will actually settle at in equilibrium is at the intersection point of the two. For purposes of our game, we can just literally use the intersection for the most part, and not worry too much about the process of it actually getting to equilibrium or all the nuances like how tariffs change things and all that BS.

How do we get these curves in DF? Well, there are a couple possible solutions to this.

Direct decision of the whole curve: You could just say "these are what the curves are" by having algorithms that entirely specify the whole curve directly. This MIGHT be simpler than other options -- if it looks like you'll get bogged down in defining the curve other ways OR it might be more complex -- if there is some simpler alternative for defining a curve with an equation or something, you probably want to go with that. You can think of this as directly defining dozens or hundreds of individual price points.

By using a few scalar values and making a curve from an equation: In this case, you might define some sort of nebulous "utility" value representing how generally useful a good is (based on anywhere to few to huge numbers of factors), and then also something like an "elasticity" value for each good, which is a value that measures basically "how much it makes sense to own multiples of this thing" or in other words "how much does demand change as price does" or in other words "the slope/curve of the line" Then using these two things plus a couple assumptions, you can fudge a passable curve dynamically generated for the trade system. Instead of defining "every" value for the line (first option), this is more like defining a line by only a point and slope, in your high school algebra class.



The more I think about it, the more I think that the DIRECT curve definition is actually the easier method for us... probably. So what would you need to do that method?
Broad Algorithm in General:
1) Set out some specific general shapes that seem reasonable for different types of goods.
2) Define necessary "control points" boundaries, asymptotes, etc. that are easy to individually calculate in the program and roughly draft out the boundaries of the desired curve.
3) Then use something like Beziers http://en.wikipedia.org/wiki/B%C3%A9zier_curve or whatever your favorite curve fitting method is, to make an actual curve based on your control values/boundaries
4) Interpolate the curve for every actual quantized unit that can be traded, and you have yourself a game-relevant demand/supply curve!
5) Figure out where they intersect for two specific trading partners.
6) Do something sort of like GoblinCookie's ordering of important goods to decide which things get traded (recalculating in between).

Starting out by discussing #1 and #2:
DEMAND:
This is very subjective. Lots of things could increase the "utility" of a good. It's just how much the buyer is driven to own it. You could want something because it will help you to survive (grain), but you could also want something just because you have a fetish for it (maybe the noble's material preferences) and you just want to roll around in a big pile of it while drooling, or you might want it to satisfy some other desire, such as revenge.  How many or few of these things we simulate is a matter of opinion and how complex you want it to be.

Each type of good should generally be assigned to one of these types of things, although again, you could always make it more or less complex.

Y axis here is in "cost" units -- if you want, imagine that 1 cost unit = one hour of dwarf labor. Here are some possible examples, though. Each curve below is not the final curve. it's just that aspect's contribution/initial input to the final curve. Any of these could/would be further tweaked up or down or by slope based on stuff like "material preferences of dwarves in the site" or political mandates or other modifiers before being used to actually determine final prices

1) Survival value -- Example FOOD. If a good is needed to sustain life, then buyers are going to be willing to pay pretty much everything they own for it up to the amount that they need to survive. Once you get past that, they will start to be willing to pay less, but the curve will have a long tail to it, representing the fact that it's always nice to have backup stockpiles of survival essentials like grain, almost to infinite quantities. The survival contribution of a good might look something like this:


*Note: the curve is not entirely vertical in drop, because the morbid reality is that a few dwarves starving to death is not actually the end of a fort, necessarily... but still, they really want to avoid that.

You could model this curve by basically defining a few individual parameters, and then fitting a curve to them.  You'd need "amount of wealth the fort has in non-survival goods" and "how much they need to survive right now" and a parameter for something like "how hoard-y/paranoid are they about having stocks on hand" and then finally, some sort of "how tasty is this type of food" parameter (quality of food doesn't matter for the left side of the curve, but does matter for the middle and right, where it will taper down much more slowly).

2) "Luxuries" I.e. generally nice things to own that aren't necessary for survival -- Example FURNITURE (also clothes, gems, etc.). "Luxury" is a very loose term here. "End user products that don't produce anything or keep you alive" if you want a more accurate term, which includes what you'd think of as actual luxuries, but also stuff like pants. Dwarves want these things, but they will never be desperate for them. Also, the curves drop down to zero or near-zero more quickly than grain, because backup stockpiles of chairs are not quite as good to have around as backup stockpiles of grain:


As you can see, this one can be vastly simpler to model. You basically just need a Y-intercept, which is going to be based on the general quality of the furniture, and then a slope "how much or little extra pieces of this furniture are" and you can make curves like this dynamically.

3) Industrial tools -- things like picks, axes, anvils, buckets, barrels, pots, maybe mechanisms? and also just metals like iron. Iron here is being abstractly conceived of as things like "chisels" and stuff that the game implies are used in stuff like carpentry shops, even though they aren't explicitly in the game. Tools are things that aren't just "nice to have" but that actually help produce additional wealth for you, thus interacting in special ways with the economy.

What I suggest for these is that the game keep track of an abstract "amount of tools" a settlement has for each individual industry. No specification. It's not necessary to actually keep track of which tools are used for which industries, but you could do that and it would be better.  The current amount of tools is used elsewhere in the supply side algorithms to act as a bottleneck for how much a settlement can ramp up production. So the major driver for demand for tools = how much below the desired level of production the settlement is, based on the prices they are seeing for things they can produce in different industries. If they are already where they want to be, they will have (almost) zero demand (still a demand, because hey, if you can get them super cheap, why not save up for future contingencies). If they are currently tool-limited, they will have fairly high demand, although the demand can't go above the rates at which it would be unprofitable, because that defeat the whole purpose:


Necessary values needed to generate this curve all shown in the graph.  You get the idea by now.

4) Building Materials -- These work exactly like industrial tools, and the curve looks sort of the same, except that these are needed to expand your population higher, so they're influenced by that instead. Instead of your desired production ramp-up driving it, it's whether or not you are currently limited by building mats for your population growth. If so, you'll pay actual reasonable values for them, or even slightly higher depending on how strapped/desperate you are (same left side slope concept as above). If you aren't limited, you'll only demand bargain basement prices from desperate sellers, otherwise you won't bother.

5) Medical? -- Kind of a special category because it's sort of like survival gear, but isn't strictly about survival. I'd draw the curve for this as something approximating the "luxuries" curve by default, except with the possibility of becoming more like the food curve temporarily IF you're in a time of war and just suffered a lot of casualties, or if your fort is ill or whatever.

6) Defense -- Just like tools and building materials, except instead of driven by desired production increase or desired population increase, it's driven by:
a) desired army/police/etc. size increase. Which is itself driven by obviously whether you're at war and whether that settlement is actually trying to build an army right now.
b) recent battles -- win or lose, a recent battle creates a directly applied need for war goods and thus if you don't have enough already, can create a shortfall that will make a little "shelf" like in the above curve under the tools section.


SUPPLY, RAW MATERIALS
This post is taking too long to write, but i'll keep my draft inside of a spoiler if you want:

Spoiler (click to show/hide)

Again, there are different broad categories, used to determine the algorithms for approximating a curve, and for how many/which parameters the code asks for. But then as with demand, the literal actual curves for each good are calculated on the fly every round of trading.  The main numbers that matter for any good are:
- Labor availability
- Tools on hand (matters more or less, depending on the good category, see below)
- Current production level
- The market situation - is it profitable to scale up? Or down? Or stay where you are?
- How much of the actual stuff is on hand

Things that are farmable (plants, livestock goods) -- These things are not only renewable, but the settlement can intentionally increase or decrease the degree to which they renew (by setting aside more or less farmland or pastures or whatever). Ramping up production is not very tool-limited, and is instead primarily geometrically limited - i.e. based on how much of the stuff you have available in stockpiles. For example, the more grain you have on hand, the more you can plant next season. The more cows you have, the more you can increase the herd in a season. If you have very little grain or cows, then you are limited in how quickly you can expand your fields or herds -- even if it would be very profitable to double the herd, you may only be able to increase by maximum 30% a season or whatever.

Things that grow/renew but not (trivially) artificially farmable (trees, fish, non-domesticated animal goods, silk, etc.) -- These things will never run out, but it is difficult or impossible to ramp up production beyond a limited point, no matter how much profit could theoretically be made from doing so. Ramping up to a certain point is easy, and just requires a moderate amount of tools, but then the industry hits a brick wall. Also, the renewability of the resource will usually decrease (slower regeneration) the higher your exploitation of it is, unlike with farms.

Non-renewable resources (primarily minerals) -- These things are finite in quantity, and also have high tool requirements for ramping up production.

SUPPLY, MANUFACTURING
haven't gotten to yet.
(needed subsection: how to decide how many raw materials to manufacture into stuff versus keeping them to sell or use them raw)

CURVES AND CURVE INTERPOLATION

SELF TRADE AND CARAVAN DETERMINATION

ORDER OF TRADE

[Scruiser]

How do we get these curves in DF? Well, there are a couple possible solutions to this.

Direct decision of the whole curve: You could just say "these are what the curves are" by having algorithms that entirely specify the whole curve directly. This MIGHT be simpler than other options -- if it looks like you'll get bogged down in defining the curve other ways OR it might be more complex -- if there is some simpler alternative for defining a curve with an equation or something, you probably want to go with that. You can think of this as directly defining dozens or hundreds of individual price points.

By using a few scalar values and making a curve from an equation: In this case, you might define some sort of nebulous "utility" value representing how generally useful a good is (based on anywhere to few to huge numbers of factors), and then also something like an "elasticity" value for each good, which is a value that measures basically "how much it makes sense to own multiples of this thing" or in other words "how much does demand change as price does" or in other words "the slope/curve of the line" Then using these two things plus a couple assumptions, you can fudge a passable curve dynamically generated for the trade system. Instead of defining "every" value for the line (first option), this is more like defining a line by only a point and slope, in your high school algebra class.

Another option: Simulate individual dwarf behaviors and motivations well enough, and give them enough options in ways of buys and selling things, and you get these curves as an "emergent" result (emergent meaning high level patterns and rules purely as the result of lower level patterns and rules).  If Toady kept integrating personality and behavior into world-gen and into more and more actions, including all economic actions, he should theoretically get some kind of supply/demand curves as an emergent result.

Supply and Demand Mechanics
Just to restate what GavJ is saying, each point on the demand curve represents a price and a quantity, the quantity is how much consumers would buy if it was that price.  Each point on the supply curve represents a price and a quantity, the quantity how much suppliers would sell for that price.  The beauty of these graphs is how much you can represent with them.  Example: for a fixed price (i.e. the government sets prices on a good), you draw a horizontal line through the graph.  You then mark where the horizontal line intersects the supply and demand curves.  You look at the quantities at each of those intersections.  The gap in quantity is either the shortage or surplus.  Its been a while since I took micro and macro economics, but basically you can represent almost any economic situation, for things as large as entire national economies (although there are issues on this scale) and for things as small as individual businesses (modifications are needed for that small a scale).  I think they work best with single markets (i.e. one particular type of good) with known characteristics.

To elaborate on my option some more.  As long as different individuals and different groups of individuals are willing to buy different quantities at different prices, and other groups/individuals/entities are willing to sell different quantities at different prices, you will get economics described by these curves as an end result.

link=topic=140342.msg5527215#msg5527215 date=1406769423]
Raw greed for wealth (gems, precious metal, coins, and high priced commodities in general), does actually sound like an interesting mechanic to me.  It would drive conflict/motivations once the basic need demand are met.  Plus it fits with the whole greedy dwarf mythology.  Rating the demands by level of importance versus affluence seems like it might work.  Like food is always high importance, but caviar and brie are luxury items that forts only prioritize as they reach a high level of wealth.

That was just my original money idea that kind of got swept under the carpet in the political argument about subsidizing caravans. 

The beauty of this system is that it simulates currency without hard-coding it into the game.  You want a society that uses stone as money, then all you do is replace coins with ANY_STONE or a particular stone if you prefer.  When it has met all it's demands, it will sell it's surplus for stone and this in turn will completely alter the economy, so that stone which at present is superabundant for everyone meaning that it is essentially untradable suddenly becomes something you can sell to caravans in vast quantities.   

If you want to get rid of money altogether, you simply set the highest affluence goods for a race/species to be finite.

Lol yeah I am actually reading your ideas before I disagree with a bunch of long winded sociopolitical-economic rants  .  I may disagree with the overall direction of your ideas, by there are definitely pieces I like/find worth discussing.  Hence why I want to get to more ideas and less abstract arguments.  I wonder how much valuation of "affluence" should be set in entity RAWs and how much calculated from personality traits.

In that case, GoblinCookie, if you aren't actually simulating industrial production itself for ANYTHING other than mining, then I'm afraid I don't really see much or any difference between our proposals?

The only things that seemed particularly meaningfully different were the lack of mention of any RAWs (which you say you do intend, though), and what appeared to be a different more detailed simulation of the production system (but apparently not). So they're almost identical... which is good I suppose.

Hence why I want to spend less time on arguing directly and more on presenting ideas.  The economics refresher was long and detailed... but you did present the idea of directly modeling complex curves to get an interesting and playable result, so I don't mind it too much.

Oh wow I didn't know that it didn't already do the reaction chains correctly. That is a good one.

I guess also along that same vein you could do the correct AMOUNTS as well as an intermediate step.

As a placeholder mechanic, the game could have priorities/rankings for certain kinds of critical goods).  All metal working civilizations have a high priority for anvils, and a requirement for an absolute minimum number for example.  All mining civs for picks (or tools associated with mining).  Etc.  Some of these could have explicit entity raws, but for others, like anvils, they could be a result of a hardcoded need associated with all civilizations that can use forges.

[GoblinCookie]

In that case, GoblinCookie, if you aren't actually simulating industrial production itself for ANYTHING other than mining, then I'm afraid I don't really see much or any difference between our proposals?

The only things that seemed particularly meaningfully different were the lack of mention of any RAWs (which you say you do intend, though), and what appeared to be a different more detailed simulation of the production system (but apparently not). So they're almost identical... which is good I suppose.

What? 

I most definately am simulating production for every good in the game, not just mining.  Mining is only different in that it extracts from a finite stockpile of natural resources and it can increase the settlements z-level range allowing both miners and other professions to use the resources at that level. 

The only real difference in the way mining production works is that other raw materials than rocks and metals (essentially) are limited in the amount that can be produced by the enviroment within the settlements radius surface territory and those underground z-levels reached.  Mined resources on the other hand are not limited but are instead finite: they run out.

The production model is the same.  The AI assigns people to do jobs according to it's demand, limited by the available stockpile of tools AND materials.  If each worker can get the neccesery tools and materials, it works at the job producing a certain fixed amount. 

This amount for finite (mined) resources is subtracted from the available amount of the resources.  For renewable resources the amount that can be produced is capped at a certain amount.  There is no other difference between the way that mining works and other resources. 

Although this will work as you say, I'm not really sure why you're describing it as a special subsection of your idea or whatever. In a proper economic model, ALL goods (or nearly/most) should have infinite demand to them. Although the demand tapers down the more of an overabundance of something you have. I.e. your 2,000th pound of rice you're willing to pay less for than your 50th pound of rice.

Settlements have a limited amount of supply and are assumed to have a limited amount of storage space though this is modelled.

But as long as a good has 1) physical space for you to store it (which is easy in DF) and 2) any conceivable possible future use whatsoever,
then it will still have continued demand. In the example of rice, dwarves would definitely have a potential use for much more rice than they eat a year: sieges.

The example you are using is already modelled by the fear mechanic.  We can make fearful settlements not only stockpile weaponry and troops but also food/drink supplies.

Therefore you should already have dozens of goods with "infinite" (asymptotically tapering) demand, and thus these would all already do what you want your money to do on their own, without having to write any special lines of code.

No we should not, it would break the system.  If a stronghold's demand for rice is NOT capped what it will do is sell everything it has in surplus in order to buy even more rice. 

The reason you do not stockpile rice is because who exactly is going to buy it?  The impoverished and starving.  The reason you are stockpile coins in because they *are* useless and only the wealthiest strongholds are going to be buying it up and they reliably have valuable things to sell.

By the way money also has that same feature in real life -- additional amounts of it are progressively worth less to you the more you have of it, just like every other commodity (this is a psychological fact). Money is just a regular commodity (outside of certain bank shenanigans, etc. which don't bother us because we don't have banks). So yeah sure I'd put an infinite (tapering) demand on ti and it would work this way, but so should a whole bunch of other things. And without any need for special code.

You stockpile money to make sure that you never actually have to sacrifice anything you actually value.  Since the system sacrifices higher demands for lower demands, it sacrifices it's infinately growing stockpile of money to make sure it promptly gets the things it actually needs. 

The infinite demands MUST be at the top of the list because mechanically the lowest infinite demands will eradicate all higher demands. 

(I'm a little bit concerned about what you MEAN by demand if not this -- infinite but tapering supply for everything except hugely bulky goods)

By demand I mean what is consumed by dwarves themselves, the militery, other settlements that are set to buy things from you and your caravan itself.  There is essentially nothing subjective or complicated about it, some demands are prioritised over others, military demands are prioritised over production, production demands are prioritised over consumer demands in general, basic consumer demands are prioritised over luxury ones, consumer demands in general are prioritised over coin-hoarding demands, everything is prioritised over trade. 

Items of a certain kind that are in excess of demand, that is considered surplus.  The computer does not object to having a surplus, it just would rather trade surplus to meet demands that it has than keep the surplus and go without what it needs.  If money if available THEN it will try and moneterise it's surplus.

Lol yeah I am actually reading your ideas before I disagree with a bunch of long winded sociopolitical-economic rants  .  I may disagree with the overall direction of your ideas, by there are definitely pieces I like/find worth discussing.  Hence why I want to get to more ideas and less abstract arguments.  I wonder how much valuation of "affluence" should be set in entity RAWs and how much calculated from personality traits.

Personality traits are irrevevent because the actual demands of individual dwarves have no economic existance for AI settlements.  They are simply a number that determines the total amount of demand of each item on the lists that have already been unlocked due to affluence.  Variations in consumer preferance if implemented should be determined by random selection within the list of items demanded by consumers at a given affluence level.

AI Settlements already have populations, that is clear from a trawl through legends mode.  In my idea the population multiplies all consumer demands by itself.  It also factors into Fear calculations because the more people you have the more you will have to stockpile and the bigger total number of soldiers are judged to be affordable.  But it does not directly affect production demands, these are determined by the total number of dwarves set to switch an occupation or presently doing it. 

Affluence is not very complex system.  Basically a settlement starts at say affluence 3 and it can go all the way up to say the magic say affluence 11 (where the infinite money demands).  Demands are checked monthly and every year affluence can go down 1 point or up 1 point. 

If we get all our existing demands, we go up an affluence level.  If we got none of the demands of our current affluence level, we go down an affluence level.  If we got some of the demands of our present affluence level but not others, we remain where we are. 

Affluence weakens a settlements position in economic terms since it creates new consumer demands, which reduce it's available surplus either because the surplus is traded away to buy these demands OR labour is transferred from producing existing surpluses to producing the desired items domestically.  Conversely lack of affluence generates surpluse for the opposite reasons, particularly when it unlocks previously reserved items like furnature. 

[GavJ]

I most definately am simulating production for every good in the game, not just mining.

I think this is just a terminology thing.
Calculating every industry the same way (except mining) but just subbing in different numbers in the same equations I do not refer to as "simulating each of them."

But the words aren't the point. My point was just that if everything except mining uses the same models, then that part is like my suggestion where everything uses the same models. In other words, we both seem to want to just use labor + tools + resources, and no additional detail for most things.

Settlements have a limited amount of supply and are assumed to have a limited amount of storage space though this is modelled.

Indeed. And coins use up finite resources to make and take up physical space... so I stand by the point that they aren't anything special or different from almost any other commodity, and that your system would work without writing any special code for them (which I will get to in my long algorithm description I'm still working on)

No we should not, it would break the system.  If a stronghold's demand for rice is NOT capped what it will do is sell everything it has in surplus in order to buy even more rice. 

No, look at the graph I posted:

See how the very high quantities have SUPER LOW prices on them?

In other words, they are willing to pay only paltry, trivial amounts of other goods for each unit of this good if they're buying that much of it.

1) This means it won't ever happen unless there's some seller who is absolutely utterly desperate to sell rice, because if not, nobody in the area will be selling it for that little value, so the trade (at that amount of volume) won't happen.
2) This means that even if there is a seller, they won't spend their whole fort's income by any means, because [whole caravan full of rice] * [ridiculously low price per unit] = modest sum of goods in return, only. 

Think of it like saying "I'm willing to buy 500 plump helmets, but only if it's for 10 dorfbucks total." Will almost never happen, but maybe someday some dude has a ton of plump helmets and is absolutely desperate, then you'll buy them. But you're only spending 10 dorfbucks. You're making out like a bandit!

If you put artificial caps on quantity like you're saying, it wouldn't be the end of the world, but it would be less realistic, because YOUR version of the dwarves would be like "Oh 500 plump helmets for 10 dorfbucks? Sorry but I have an abstract brick wall coded into me at 125." which is not what would happen. Also, you're adding complexity to modders and making it harder to balance by having to decide what all those caps are, versus letting the asymptote deal with it automatically in almost every case.

The infinite demands MUST be at the top of the list because mechanically the lowest infinite demands will eradicate all higher demands. 

You're just misunderstanding how demand works. Please see my previous long post where I talk about supply and demand curves, or look up the relevant articles on wikipedia or economics teaching websites - it's easy to get the hang of in an hour or so if you aren't familiar.

What you're describing only makes sense if you unrealistically think of demand as flat, not if you think of it as a proper curve. I don't know how better to explain it than the post I put above, in any efficient amount of time, so again I just suggest wikipedia, etc.

By demand I mean what is consumed by dwarves themselves, the militery, other settlements that are set to buy things from you and your caravan itself.  There is essentially nothing subjective or complicated about it, some demands are prioritised over others, military demands are prioritised over production, production demands are prioritised over consumer demands in general, basic consumer demands are prioritised over luxury ones, consumer demands in general are prioritised over coin-hoarding demands, everything is prioritised over trade.

Yes, that's a problem. You can't do demand that simply and have a realistic economy work out. Certain goods do have priorities (like Maslow's hierarchy of needs sort of thing), like you say, for sure. But you have to FOLD THOSE INTO the normal demand curves, not treat them as a special thing of their own.

Well you could do it either way, but fiddling manually is just inefficient and weird IMO. And I think it's going to feel stilted and broken. I'm having trouble describing this one, but sooner or later, I'll get to this same part of the algorithm I'm describing, and hopefully it will be clear what I mean and we can discuss further.

I would still intend to have an order of trade, but it would be less important and not the entire driving force. blergh... yeah just will explain more later.

[GoblinCookie]

I think this is just a terminology thing.
Calculating every industry the same way (except mining) but just subbing in different numbers in the same equations I do not refer to as "simulating each of them."

But the words aren't the point. My point was just that if everything except mining uses the same models, then that part is like my suggestion where everything uses the same models. In other words, we both seem to want to just use labor + tools + resources, and no additional detail for most things.

They use the same model because surprisingly for everything in the game but mining the same model will actually work.  You have seemingly mistaken the slightly different system that mining uses for implying that I had not got a model for everything else. 

Indeed. And coins use up finite resources to make and take up physical space... so I stand by the point that they aren't anything special or different from almost any other commodity, and that your system would work without writing any special code for them (which I will get to in my long algorithm description I'm still working on)

The would take up less physical space per value, including on caravans than anything else in the game.  But as I said, if you want something else to take it's place you can, whether it makes actual economic sense or not. 

No, look at the graph I posted:

See how the very high quantities have SUPER LOW prices on them?

Graphs give me vertigo.

That is utterly senseless even as a basic economic concept let along a gameplay mechanics.  You have the item, they want the item.  It matters not whether you have a small amount of it or a lot of it, the only thing that matters here is how much they want what you want and how much you want what they want AND most importantly who else can offer them what they want. 

If you have stockpiled a ridiculous amount of of the rice available, what you are doing is cornering the rice market.  You are acting as a rice monopolist essentially.  You are able to give the stuff away for free, this means competing with you is madness, that in turn makes you into a monopolist; that means you can actually charge high prices for a superabundant but monopolised good. 

You are *able* to give it away for free, but you would rather sell it for a high fee.  The first fact eliminates competition, ensuring that you are able to achieve the second end.

Think of it like saying "I'm willing to buy 500 plump helmets, but only if it's for 10 dorfbucks total." Will almost never happen, but maybe someday some dude has a ton of plump helmets and is absolutely desperate, then you'll buy them. But you're only spending 10 dorfbucks. You're making out like a bandit!

Out comes the fictitious imaginary currency again.  It amazes me how you cannot actually model your economic system based upon the actual mechanics of the game as it stands.  You are not trading dwarfbucks, you are trading valuable and scarce items that took labour to produce and additional labour to transport.

If you get 500 plump helmets, what would you do with them?  Do enough people want to buy plump helmets (something almost everyone can produce) to actually exchange them for valuable items?  But if we were living in a world that was nearing Malthusian point, all but the most soil-abundant settlements would start to demand plump helmets BUT since you are still not the only person in the world that has soil you still cannot exactly dictate prices at whim. 

If you put artificial caps on quantity like you're saying, it wouldn't be the end of the world, but it would be less realistic, because YOUR version of the dwarves would be like "Oh 500 plump helmets for 10 dorfbucks? Sorry but I have an abstract brick wall coded into me at 125." which is not what would happen. Also, you're adding complexity to modders and making it harder to balance by having to decide what all those caps are, versus letting the asymptote deal with it automatically in almost every case.

The system I am using is quite simple AND realistic.  There is no abstract brick wall here, the caravan refuses to buy more than 125 plump helmets because that is the combined amount that.

1. It uses.
2. It's owner demands.
3. It can sell to others. 

The caps are not artificial.  They are naturally emerging based upon the economic situation in the game.  Nobody will buy things they cannot sell or use themselves because they take up space in the caravan or the stockpile that could be used for things they *could* use or sell and valuable things had to be traded to get them (since if you are willing to buy them, they are valuable). 

In order to get those extra goods they had to sell other items that are not in such abundance.  It is totally unrealistic that you can presently sell a limitless amout of non-scarce goods far in excess of the goods that the traders can actually hope to use or sell. 

You're just misunderstanding how demand works. Please see my previous long post where I talk about supply and demand curves, or look up the relevant articles on wikipedia or economics teaching websites - it's easy to get the hang of in an hour or so if you aren't familiar.

I understand quite well how my own economic system works.  I was merely informing you as to what will happen if we give items lower in affluence level than the top infinite demand in my system.  It will happen because of the way the AIs are programmed to make their economic decisions. 

What you're describing only makes sense if you unrealistically think of demand as flat, not if you think of it as a proper curve. I don't know how better to explain it than the post I put above, in any efficient amount of time, so again I just suggest wikipedia, etc.

I thought I had already modelled non-flat demands through the whole concept of affluence levels. 

Yes, that's a problem. You can't do demand that simply and have a realistic economy work out. Certain goods do have priorities (like Maslow's hierarchy of needs sort of thing), like you say, for sure. But you have to FOLD THOSE INTO the normal demand curves, not treat them as a special thing of their own.

Yes things really are that simple.  If your textbooks say otherwise, simply burn your textbooks. 

Both in the game and in reality people have finite demands for various things.  As they get the basic demands, they develop demands that are more and more rarified, as well as usually more and more costly.  But despite this being the case, overall certain people tend to consistantly manage to hoard endless millions of <insert currency name here>. 

This establishes quite clearly that while all other demands are finite, a certain demand is not. That demand is the demand for money.  The difference between dwarf fortress and reality is only that dwarves do not personally own and use money, this does not stop the same situation from developing at the fortress level. 

Well you could do it either way, but fiddling manually is just inefficient and weird IMO. And I think it's going to feel stilted and broken. I'm having trouble describing this one, but sooner or later, I'll get to this same part of the algorithm I'm describing, and hopefully it will be clear what I mean and we can discuss further.

I would still intend to have an order of trade, but it would be less important and not the entire driving force. blergh... yeah just will explain more later.

I do not get the idea about fiddling manually from?  I do sense that yet another strawmen is in the process of being born.

[GavJ]

That is utterly senseless even as a basic economic concept let along a gameplay mechanics.  You have the item, they want the item.  It matters not whether you have a small amount of it or a lot of it, the only thing that matters here is how much they want what you want and how much you want what they want

If you're just going to ignore/deny concepts that have been around and nearly universally accepted and successfully applied to things just like this since the 17th century, on par with the the level of knowledge of printing presses or telescopes, then there really isn't much else productive to discuss with you on the topic.

So I'm going to stop spending time typing about that stuff and focus now only on clarifying technical details of the algorithms. There seems to be one important comment in this post in that regard:

Out comes the fictitious imaginary currency again.

To be clear: "dorfbucks" are not currency or intended to be interpreted as such. "Dorfbucks" is just the traditional, community-given name for the existing value system. 1 Dorfbuck = a unit of generic value, which all goods have some amount of, and any good can fill in for. Currently, this value is just defined directly in the raws for materials. In the proposed economics model, though, this value would be defined by the whole demand/supply dynamic curves model, etc.

[GoblinCookie]

If you're just going to ignore/deny concepts that have been around and nearly universally accepted and successfully applied to things just like this since the 17th century, on par with the the level of knowledge of printing presses or telescopes, then there really isn't much else productive to discuss with you on the topic.

Falsehoods do not become true just because they have decieved a lot of people for a very long time.  People thought the sun went round the earth for a very long time as well.

You want us to automatically reset prices artificially according to supply and demand without actually modelling the actual process that makes this happen.  You want the implementation of some kind of economic mysticism, you want the process to happen automatically without having a realistic mechanism that actually models this. 

In reality it is the general supply or shortage that affects prices due to increasing or reducing competition, not the amount actually in the hands of the entity that is trading. While this can be modelled easily given a 'box' where everyone can trade with eachother for equal cost, the lack of such natural boxes in the game world means that it will not be able to model the 'local' competitive availability without having to have everyone have to individually calculate the amount of competition for every single settlement based upon what everyone else *could* be doing and the relative costs of doing so etc.

To be clear: "dorfbucks" are not currency or intended to be interpreted as such. "Dorfbucks" is just the traditional, community-given name for the existing value system. 1 Dorfbuck = a unit of generic value, which all goods have some amount of, and any good can fill in for. Currently, this value is just defined directly in the raws for materials. In the proposed economics model, though, this value would be defined by the whole demand/supply dynamic curves model, etc.

In my model the value of objects has a very limited role because AIs set up trades based upon their supply and demand directly.  It will simply not buy in excess of it's demands and it will not sell what it cannot afford to lose unless to get things it needs more desperately.

[GavJ]

You can't possibly know that a theory is wrong when you don't know the theory yet.

the amount actually in the hands of the entity that is trading

Quantity in a supply/demand curve is not quantity owned. It's quantity changing hands in a given instance. Which should actually be intuitive:

Imagine you live alone in a house with everything except chairs. Not a chair anywhere. You're pretty desperate for a place to sit down, so you go shopping. Turns out chairs are a little pricier than you can afford to fill up your whole house with them (all the dining room ones, office, living room, etc.). Maybe they go for $100 but you can't furnish your whole house unless they are $70 or cheaper. BUT at that same price, you CAN afford and are in fact willing to pay for one chair, because you need a place to sit

As the amount of chairs being bought goes up, though, you become less willing to pay as much per unit. Because the first chair has more value to you than subsequent chairs. Why? Because you can only sit in one chair at a time. Additional ones only provide lesser value - flexibility in your sitting, entertainment which happens less often, etc. And in fact the more chairs you add, the less valuable each next one becomes, because a dinner party with 8 people is probably less common than a dinner party with 4, and the flexibility for you that it added by having multiple chairs in a room is not much.

HOWEVER, chairs still always have value. Even to you potentially for rare usages, but also just because you can resell them. So if you were to find a seller somewhere that was offering 50,000 high quality chairs for $0.01 each, even though they're just as nice of chairs as the ones further up, you should buy them.  So what if you have to rent out storage spaces? So what if you have to quit your job to find ways to sell them all? That's all fine, because if you can manage to sell them for even half of what that guy in the first part of the story was, you will make $2,499,500 in profit, easily paying your storage and a much higher wage to you probably than your previous job. Hell, for that much profit, you could even afford to rent out a special store space on main street with a big sign saying "CHAIR WORLD!" to get higher prices and easier sales.

So basically you always have no actual cap on how much of something you're willing to buy, IF the price is insanely low enough.

Hence, a graph like this:


At 1 chair bought, you're willing to pay up to $120 (doesn't mean you will. At $100 you'll take it too!)
At 10 chairs (the amount to furnish your whole house), you are only willing to pay $70 or less per chair.
At 50,000 chairs, you're unwilling to buy unless the price is a penny or less per chair. Sure, even twice this price could still make you a millionaire, but the stress of quitting your job and the uncertainty etc. factors in so that you're only willing to take the plunge if it gets more tempting than that, i.e. 2.5 million potential profit.

Gray = area under each rectangle = actual amount of money changing hands for that many chairs. Notice, importantly, that none of those numbers are bankrupting the guy. None of them are "all the stuff in his entire fort."

And after any given transaction, the curve is going to change to reflect the differing values chairs now have for you.

[GoblinCookie]

You can't possibly know that a theory is wrong when you don't know the theory yet.

Quite. I just do not trust economists or economic theories because they masquerade as objective science what are actually deductions from psychological/social/political assumptions and often overt or covert agendas. 

Quantity in a supply/demand curve is not quantity owned. It's quantity changing hands in a given instance. Which should actually be intuitive:

Imagine you live alone in a house with everything except chairs. Not a chair anywhere. You're pretty desperate for a place to sit down, so you go shopping. Turns out chairs are a little pricier than you can afford to fill up your whole house with them (all the dining room ones, office, living room, etc.). Maybe they go for $100 but you can't furnish your whole house unless they are $70 or cheaper. BUT at that same price, you CAN afford and are in fact willing to pay for one chair, because you need a place to sit

In dwarf fortress barter economy terms, you are probably not trading in cash, you are trading in surplus.  Unless you have a surplus, you will only be able to buy chairs IF you sacrifice something else.  What you will rationally do is sacrifice pretty gemstones and jewelry, this less useful than chairs in order to buy what you cannot get with your surplus.

For your $100 chair set, you will sacrifice your $70 surplus and then you will sacrifice $30 of jewelry. 

As the amount of chairs being bought goes up, though, you become less willing to pay as much per unit. Because the first chair has more value to you than subsequent chairs. Why? Because you can only sit in one chair at a time. Additional ones only provide lesser value - flexibility in your sitting, entertainment which happens less often, etc. And in fact the more chairs you add, the less valuable each next one becomes, because a dinner party with 8 people is probably less common than a dinner party with 4, and the flexibility for you that it added by having multiple chairs in a room is not much.

This is already modelled through my affluence system.  Your 4 chairs for normal use is affluence say 3.  When you get all your chairs (and everything else) you will upgrade to affluence 4 where we could need double the number of chairs. 

Once you have reached the maximum number of chairs you could possibly want, we have reached a point where chairs have suddenly no actual value.  They are of negative (?) value now, they are something you will spend resources on caravans in order to get rid of if you can get absolutely anything in return that actually has value. 

Or you could give them away as charity, use them to bribe the local goblins to leave you alone, send them off to the capital in a bid to get yourself promoted to duke or so on.

HOWEVER, chairs still always have value. Even to you potentially for rare usages, but also just because you can resell them. So if you were to find a seller somewhere that was offering 50,000 high quality chairs for $0.01 each, even though they're just as nice of chairs as the ones further up, you should buy them.  So what if you have to rent out storage spaces? So what if you have to quit your job to find ways to sell them all? That's all fine, because if you can manage to sell them for even half of what that guy in the first part of the story was, you will make $2,499,500 in profit, easily paying your storage and a much higher wage to you probably than your previous job. Hell, for that much profit, you could even afford to rent out a special store space on main street with a big sign saying "CHAIR WORLD!" to get higher prices and easier sales.

They have value only as money (that is apex affluence infinite demand item).  The problem is that noone else is using chairs as money, unless chairs are a recognised (and legal) currency then chairs as described above become worthless the moment your exceed the maximum possible usage for chairs.  If you cannot get rid of them you will probably start to chop them down into wood in order to burn to make charcoal to forge metals or something.   

The wood they are made of is now literally worth more than the chairs are.  I have already taken into account the stockpiling of chairs in order to sell them to others, the demands of other strongholds you already have trade relations with are considered demand but have the lowest priority. 

The way this would work in practice does give me vertigo even though it is not complicated as a concept, but it allows for the arising of trade hubs rather nicely. 

So basically you always have no actual cap on how much of something you're willing to buy, IF the price is insanely low enough.

Hence, a graph like this:

This is where your economic theories err. 

If you are stockpiling chairs that you neither use nor intend to sell because they are a store of value, you do buy items that are of a low price, because those you are trading are still trying to make a profit.  The know the items are worthless, that they are selling them to you for something, however little means that what you are giving them is worth more than what you are getting. 

The only reason that you would actually collect infinite chairs is if you were using chairs actually as money (abstract store of value).  I have already included that option, to make chairs money all you have is to add ChairxInfinity to the top affluence level (but not anything below as it will break the game).  This of course should be illegal and the seriously left-wing governments of dwarf universe world would doublessly have no compunctions in making this so. 

The reason is that using useful objects as money is very destructive because it creates a shortage of whatever it is, meaning it becomes artificially scarce so that those you wish to make actual use of them and are not wealthy get bidded out by wealthy hoarders.  To delve into politics, one of the main economic problems in the world is that money is not only a well infinite money demand but also a 'production item' (because people are paid in it). 

To model real-world economics one would make money a production demand and keep it as an infinite demand apex affluence good. Then we would find that loads of people cannot afford to make their dwarves work, resulting in mass-unemployment, collapse of production, economic ruin and instability (all sorts of real-life problems). 

At 1 chair bought, you're willing to pay up to $120 (doesn't mean you will. At $100 you'll take it too!)
At 10 chairs (the amount to furnish your whole house), you are only willing to pay $70 or less per chair.
At 50,000 chairs, you're unwilling to buy unless the price is a penny or less per chair. Sure, even twice this price could still make you a millionaire, but the stress of quitting your job and the uncertainty etc. factors in so that you're only willing to take the plunge if it gets more tempting than that, i.e. 2.5 million potential profit.

Gray = area under each rectangle = actual amount of money changing hands for that many chairs. Notice, importantly, that none of those numbers are bankrupting the guy. None of them are "all the stuff in his entire fort."

And after any given transaction, the curve is going to change to reflect the differing values chairs now have for you.

In dwarf fortress terms all chairs above 10 (maximum affluence level) would be worthless.  As already explained, all that happens if you start selling chairs ridiculously cheaply (which nobody would do anyway) is that people will start to chop the chair up and make charcoal out of the lot of them. 

Of course it would never actually happen in my world economic system. It is unlikely that anyone would actually trade chairs at all, they would simply import the wood (and probably only if they have no trees) to make their own chairs.  That is kind of an improvement on the present setup in my opinion.

Instead of having to trawl through hundreds of items to get to one's you want, you would only be given a finite set of items in a finite quantity based initially on what the AI calculates your demand should be (what it would be if you were an AI) based upon what your fear rating would be, your affluence would be, your production would be based upon your professions (and so on). 

You would however be able to sign trade contracts with specific AIs in order to get goods you specifically want.  This would adjust your percieved demand to the AIs in general, so it quite possible the AI that you originally contracted with might well be outmanuevred by another AI, which could be used as the basis for political trouble. 

[GavJ]

In dwarf fortress barter economy terms, you are probably not trading in cash, you are trading in surplus.  Unless you have a surplus, you will only be able to buy chairs IF you sacrifice something else.  What you will rationally do is sacrifice pretty gemstones and jewelry, this less useful than chairs in order to buy what you cannot get with your surplus.

I do agree that in general, you should consider selling the things of yours that are least useful to you first. However:

1) Gemstones are not the most useless things, because they do have a use in game. They make transparent windows that are often very useful for traps and sentries and other things (sea serpent breeding, etc.)
2) Other things actually do have zero use in game. Like I dunno, limestone flutes. So if you're going purely by order of usefulness, you should try to sell instruments and things with no use prior to gems.
3) However, most importantly, instruments are also not useful to the other guy. Things that have literally zero use in game should have zero value for trading. So you shouldn't be able to sell them in exchange for anything, because your trading partners should refuse to take them, due to having zero demand for them.

Sure you could adjust your raws or whatever to make NPCs consume useless things, but that's basically cheating, and is not what the default values should be. For a realistic and challenging experience, a player should set the NPC consumptions to be basically equal to his own consumptions, and only exist for things that actually have in game uses. I.e. they should demand zero flutes, incredibly few gems, and so on (represented by very short, very steep demand curves)

In other words, to get anything in exchange for your stuff, you should ALWAYS have to make some actual useful sacrifices. Although you would do so in order of least sacrifice, as you point out, yes.

This is already modelled through my affluence system.  Your 4 chairs for normal use is affluence say 3.  When you get all your chairs (and everything else) you will upgrade to affluence 4 where we could need double the number of chairs.

Except instead of using a weird system you made up in half an hour and that is overly crudely measuring "tiers" of numbers, we can just as easily use the actual, professional system for doing the exact same thing that has been in use for hundreds of years, offers greater precision down to the individual numbers of items, is already known to work, and is able to model multiple different motivations in one curve for things other than affluence (the same curve can simulate both affluence AND wililngness to make business investments, etc., like my example portrayed -- the guy buying 50,000 chairs isn't doing it because he's Uncle moneybags and is showing off how many chairs he has. It switches to a business venture at that point, yet can still be captured with one system.)

The beauty of supply/demand curves, as Scruiser pointed out, if that one curve can capture a dozen different motivations and plans, or technically infinitely many, all in one standardized system. Affluence, by comparison, captures one motivation: showing off affluence. And it does so weirdly at that, with these sudden thresholds that are not realistic "Oh hey we own 5 chairs now. We are now officially level 3 rich! And as of this very second suddenly desire exaclty 17 chairs whereas 5 minutes ago we never considered the possibility of desiring more than 5!!" Wtf? Nobody thinks like that.

Once you have reached the maximum number of chairs you could possibly want, we have reached a point where chairs have suddenly no actual value.

So if you had as many chairs as you wanted already for your level of "affluence," and I drove up to your house, and I offered you a legitimate Elizabethan royal hand carved English chair for $5, you're saying you wouldn't buy it, because you've reached your limit of chairs and chairs have no value for you?  Yeah, that makes tons of realistic sense, dude.

They know the [chairs] are worthless

No! They AREN'T worthless - they're perfectly high quality, useful chairs. I was very careful to specify that multiple times in my post.

The guy selling them to me for $0.01 is selling them way under what they are worth.  There are lots of reasons people do this occasionally. For example, his settlement might be starving to death and they will all literally die if he doesn't get enough grain during this trade he's doing with you. So even though he knows he is losing a ton of value on chair, he is forced to do so for precious precious grain (to him). In fact, the methods described by I think either of our methods (certainly mine I am sure) would allow this exact situation to happen in game.

There could be some dudes a couple miles further away who will happily buy all those chairs at full price a day later, but he just couldn't wait that long to make the further trip. HAD to get his grain THEN, etc. etc.

You not buying them in those circumstances would be ridiculously wasting a huge opportunity.




edit: it occurs to me that I could simply program the system I want, prior to explaining it all. Since it doesn't actually involve much GUI or things frmo the game that would be difficult to bandaid for a test.

[GoblinCookie]

I do agree that in general, you should consider selling the things of yours that are least useful to you first. However:

1) Gemstones are not the most useless things, because they do have a use in game. They make transparent windows that are often very useful for traps and sentries and other things (sea serpent breeding, etc.)
2) Other things actually do have zero use in game. Like I dunno, limestone flutes. So if you're going purely by order of usefulness, you should try to sell instruments and things with no use prior to gems.
3) However, most importantly, instruments are also not useful to the other guy. Things that have literally zero use in game should have zero value for trading. So you shouldn't be able to sell them in exchange for anything, because your trading partners should refuse to take them, due to having zero demand for them.

1.  That is why gemstones are ranked below money in the list of thinks.  Raw gold and silver is also ranked below money as well.  But pretty much all furnature is ranked below gemstones.
2.  Correct, limestone flutes would be unsellable, along with a lot of other things (such as the vast majority of ordinery stone). 
3. They do not have zero value, they simply are not sellable because caravans refuse to buy in excess of demand which in this case is 0.

Sure you could adjust your raws or whatever to make NPCs consume useless things, but that's basically cheating, and is not what the default values should be. For a realistic and challenging experience, a player should set the NPC consumptions to be basically equal to his own consumptions, and only exist for things that actually have in game uses. I.e. they should demand zero flutes, incredibly few gems, and so on (represented by very short, very steep demand curves)

That is basically my idea. 

In other words, to get anything in exchange for your stuff, you should ALWAYS have to make some actual useful sacrifices. Although you would do so in order of least sacrifice, as you point out, yes.

Actually there is no sacrifice at all in getting rid of surplus items because these actually have no value to the person who has them.  All sacrifice is irrational because it means you are selling something worthless (not in demand) for less valuable goods than you could acquire elsewhere.

Except instead of using a weird system you made up in half an hour and that is overly crudely measuring "tiers" of numbers, we can just as easily use the actual, professional system for doing the exact same thing that has been in use for hundreds of years, offers greater precision down to the individual numbers of items, is already known to work, and is able to model multiple different motivations in one curve for things other than affluence (the same curve can simulate both affluence AND wililngness to make business investments, etc., like my example portrayed -- the guy buying 50,000 chairs isn't doing it because he's Uncle moneybags and is showing off how many chairs he has. It switches to a business venture at that point, yet can still be captured with one system.)

Your person buying 50,000 chairs for a 'buisness venture' is basically buying them in order to sell them on.  If he cannot sell 50,000 chairs, he will not buy 50,000 chairs. In game terms he genuinely does demand chairs, he demands them as stock for his furnature mega-store and again they are only valuable in so far as he can sell them on.  The number of chairs that have value is limited by the available customer base. 

However since people are practically giving away the chairs for free, even such a venture is of dubious value. It thus determined that the chairs are worthless and they are to be rendered down into their componant wood to make charcoal. 

The beauty of supply/demand curves, as Scruiser pointed out, if that one curve can capture a dozen different motivations and plans, or technically infinitely many, all in one standardized system. Affluence, by comparison, captures one motivation: showing off affluence. And it does so weirdly at that, with these sudden thresholds that are not realistic "Oh hey we own 5 chairs now. We are now officially level 3 rich! And as of this very second suddenly desire exaclty 17 chairs whereas 5 minutes ago we never considered the possibility of desiring more than 5!!" Wtf? Nobody thinks like that.

Everyone thinks like that;that is actually how consumer demand works.  First I look to acquire basic survival goods, then crude functional goods and finally 'artistic' goods but if they have no surplus a person will sacrifice the higher goods for the lower.  Of course it is not as tidy as having an actual number but that is the way computers work inherantly.

I have already demonstrated there is no demand curve in real-life or in the game.  There is only a demand cliff, either something has a value for some reason or it has no value for anything except being rendered down into whatever componant parts may actually have some value. 

So if you had as many chairs as you wanted already for your level of "affluence," and I drove up to your house, and I offered you a legitimate Elizabethan royal hand carved English chair for $5, you're saying you wouldn't buy it, because you've reached your limit of chairs and chairs have no value for you?  Yeah, that makes tons of realistic sense, dude.

That is a deceptive argument because you have just offered me a higher affluence item than the regular chair.  If I already had a house full of such chairs then I would actually be stupid to accept the offer at any price.

You could sell them on for a higher price and hope the guy has not completely managed to flood the market with said chairs at the lower price I bought them for (or worse), but that would mean that the chairs are actually in-demand, not exactly evidence of your demand curve then but rather of my demand cliff. 

The guy selling them to me for $0.01 is selling them way under what they are worth.  There are lots of reasons people do this occasionally. For example, his settlement might be starving to death and they will all literally die if he doesn't get enough grain during this trade he's doing with you. So even though he knows he is losing a ton of value on chair, he is forced to do so for precious precious grain (to him). In fact, the methods described by I think either of our methods (certainly mine I am sure) would allow this exact situation to happen in game.

There could be some dudes a couple miles further away who will happily buy all those chairs at full price a day later, but he just couldn't wait that long to make the further trip. HAD to get his grain THEN, etc. etc.

You not buying them in those circumstances would be ridiculously wasting a huge opportunity.

The grain you are selling is surplus and thus worthless to you. The actual economic decision here is whether to sell your grain to the starving settlement or a settlement that has something to offer you that is valuable.  If the chairs are also worthless to you then the grain is worthless to you also but you could otherwise have sold the grain for something valuable and thus actually gained something by the exchange of your limited surplus. 

If you can sell the chair on to someone else, well that is already part of the system it means that the chair is actually worth something because you can sell it on.  But you cannot acquire anything of value to you (or someone else) by selling to the third guy, the chair is still worthless and no trade will occur at any price.  The object is either IN DEMAND or it is NOT IN DEMAND.

Inherantly then value is binary, an object has value or it has no value relative to a quantity of such items.  What leads to illusions such as the demand curve is simply the introduction of money into the system.  Because money is of infinite demand (have more money will always be valuable) and worthless objects to you are still exchangable for money, it 'appears' that worthless objects have actual value because as long as they can potentially be exchanged for a quantity of money greater than the amount of money handed over OR the amount of money than the bartered objects could have been exchanged for. 

Of course the funny thing is that the objects are not valuable, only the money is valuable.  The objects may have thrown themselves over the demand cliff long ago, but money has no demand cliff. This is shown by how you cannot express the concept except in monetery terms.

[Scruiser]

That is utterly senseless even as a basic economic concept let along a gameplay mechanics.  You have the item, they want the item.  It matters not whether you have a small amount of it or a lot of it, the only thing that matters here is how much they want what you want and how much you want what they want

If you're just going to ignore/deny concepts that have been around and nearly universally accepted and successfully applied to things just like this since the 17th century, on par with the the level of knowledge of printing presses or telescopes, then there really isn't much else productive to discuss with you on the topic.

So I'm going to stop spending time typing about that stuff and focus now only on clarifying technical details of the algorithms.

/start rant
Pretty much the conclusion I reached a few pages back.  GoblinCookie, economist don't have everything right, but it is not the "argument from authority" to say that the field of study that has worked on issues for over a hundred years might have some accurate things to say about those issues.  At the very least, you will need to give more justification for why what you have thought up in your spare time is so much more useful, accurate and true than what many other people have spent a long time working on.  In the future, could you just ignore my posts based around economics, and I'll ignore your posts that go against economics and hopefully we can avoid cluttering threads with long arguments that will go nowhere?  I really don't see that as productive for generating good suggestions.
/end rant

I am thinking of starting a new thread focused solely on economic ideas that can be done in the next set of major feature updates.  Small changes and small feature additions that will be playable and interesting, while laying the ground work for further improvements.  I think focusing on the short term, we can exchange some good ideas without getting into these wall of text arguments.
What do you two think? (Since it seems to be only GavJ and GoblinCookie left on this thread)